pytorchjob-generator: add successTTLDuration (#129)

Author: tardieu

tools/pytorchjob-generator/chart/README.md

@@ -71,7 +71,8 @@ customize the Jobs generated by the tool.
 | failureGracePeriodDuration | string | The AppWrapper defaults will be used | Customize the failureGracePeriod; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/ |
 | retryPausePeriodDuration | string | The AppWrapper defaults will be used | Customize the retryPausePeriod; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/ |
 | retryLimit | integer | The AppWrapper defaults will be used | Customize the retryLimit; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/ |
-| forcefulDeletionGracePeriodDuration | string | The AppWrapper defaults will be used | Customize the forcefulDelectionGracePeriod; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/ |
+| forcefulDeletionGracePeriodDuration | string | The AppWrapper defaults will be used | Customize the forcefulDeletionGracePeriod; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/ |
 | deletionOnFailureGracePeriodDuration | string | The AppWrapper defaults will be used | Customize the deletionOnFailureGracePeriod; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/ |
-| restartPolicy | string | `"Never"` | Set Kubernertes policy for restarting failed containers "in place" (without restarting the Pod). |
+| successTTLDuration | string | The AppWrapper defaults will be used | Customize the successTTL; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/ |
+| restartPolicy | string | `"Never"` | Set Kubernetes policy for restarting failed containers "in place" (without restarting the Pod). |
 | terminationGracePeriodSeconds | integer | Kubernetes's default value is used | Set a non-default pod termination grace period (in seconds). |

tools/pytorchjob-generator/chart/templates/appwrapper.yaml

@@ -78,6 +78,9 @@ metadata:
         {{- if .Values.deletionOnFailureGracePeriodDuration }}
         workload.codeflare.dev.appwrapper/deletionOnFailureGracePeriodDuration: "{{ .Values.deletionOnFailureGracePeriodDuration }}"
         {{- end }}
+        {{- if .Values.successTTLDuration }}
+        workload.codeflare.dev.appwrapper/successTTLDuration: "{{ .Values.successTTLDuration }}"
+        {{- end }}
     {{- if or .Values.queueName .Values.customLabels }}
     labels:
         {{- if .Values.queueName }}

tools/pytorchjob-generator/chart/tests/helloworld_test.yaml

@@ -186,6 +186,7 @@ tests:
     retryLimit: 42
     forcefulDeletionGracePeriodDuration: "19s"
     deletionOnFailureGracePeriodDuration: "2s"
+    successTTLDuration: "600s"
   asserts:
   - isSubset:
       path: metadata.annotations
@@ -197,6 +198,7 @@ tests:
         workload.codeflare.dev.appwrapper/retryLimit: "42"
         workload.codeflare.dev.appwrapper/forcefulDeletionGracePeriodDuration: "19s"
         workload.codeflare.dev.appwrapper/deletionOnFailureGracePeriodDuration: "2s"
+        workload.codeflare.dev.appwrapper/successTTLDuration: "600s"
 
 - it: Setting integer fault tolerance annotation to 0
   set:

tools/pytorchjob-generator/chart/values.schema.json

@@ -153,6 +153,10 @@
         "deletionOnFailureGracePeriodDuration" : { "oneOf" : [
             { "type": "null" },
             { "$ref": "#/$defs/duration" }
+        ]},
+        "successTTLDuration" : { "oneOf" : [
+            { "type": "null" },
+            { "$ref": "#/$defs/duration" }
         ]}
     },
 

tools/pytorchjob-generator/chart/values.yaml

@@ -255,7 +255,7 @@ retryPausePeriodDuration:
 # @default -- The AppWrapper defaults will be used
 retryLimit:
 
-# -- (string) Customize the forcefulDelectionGracePeriod; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/
+# -- (string) Customize the forcefulDeletionGracePeriod; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/
 # @section -- Fault Tolerance
 # @default -- The AppWrapper defaults will be used
 forcefulDeletionGracePeriodDuration:
@@ -265,7 +265,12 @@ forcefulDeletionGracePeriodDuration:
 # @default -- The AppWrapper defaults will be used
 deletionOnFailureGracePeriodDuration:
 
-# -- (string) Set Kubernertes policy for restarting failed containers "in place" (without restarting the Pod).
+# -- (string) Customize the successTTL; see https://project-codeflare.github.io/appwrapper/arch-fault-tolerance/
+# @section -- Fault Tolerance
+# @default -- The AppWrapper defaults will be used
+successTTLDuration:
+
+# -- (string) Set Kubernetes policy for restarting failed containers "in place" (without restarting the Pod).
 # @section -- Fault Tolerance
 restartPolicy: "Never"